Oxytocin is important for many aspects of social cognition. However, it is far from understood how oxytocin acts in the brain to have its effects on social perception, learning and the formation of long-term attachments. The monogamous prairie vole (Microtus ochrogaster) is a premier animal model for studying mechanisms of attachment, and oxytocin acting in various reward-related brain regions is essential for prairie voles to form pair bonds. Yet exactly how oxytocin facilitates pair bonding by modulating the underlying neural circuitry during social interactions to form bonds is unknown, pointing to a gap in our knowledge linking neurochemical to neural circuit mechanisms of attachment. Our long-term goal is to elucidate how oxytocin modulates reward and sensory systems underlying social information processing and learning. We focus here on a key oxytocin receptor-rich node at the interface between these systems, the nucleus accumbens, which receives inputs from other oxytocin receptor-dense areas, the medial prefrontal cortex (Aim 1) and the basolateral amygdala (Aim 2). Our objective here is to determine whether manipulating the oxytocin system to impair or enhance pair bonding affects nucleus accumbens’ functional connectivity with its mPFC and BLA inputs. Our central hypothesis is that oxytocin normally acts to improve communication from reward and cue processing areas to local NAc circuits that integrate these channels of information during specific social interactions, helping to reinforce the ability of partner signals to elicit affiliative behavior. We validate this hypothesis using both loss-of-function and gain-of-function experimental designs, as well as optogenetics to test causality. The rationale for our proposal is that, once we know how oxytocin affects neural circuitry between brain areas to facilitate the formation of a selective attachment, we can begin to elucidate the molecular mechanisms for plasticity within these circuits.